Autonomous Recording Units Research Articles

Essential steps for establishing a large‐scale passive acoustic monitoring for an elusive forest bird species: the Eurasian Woodcock (Scolopax rusticola)

In light of global biodiversity loss, comprehensive monitoring of species and biodiversity is fundamental for effective conservation management. Technology‐driven approaches, particularly in the field of bioacoustics, have gained importance, with autonomous recording units (ARUs) such as the AudioMoth being a notable example. However, a systematic application of this technology in large‐scale passive acoustic monitoring (PAM) schemes is challenging because of a lack of species‐specific calibration techniques and information on survey design. Using Eurasian Woodcock Scolopax rusticola as our focal species, we present application‐oriented guidelines for testing and implementing a terrestrial, single‐species PAM scheme. In this process, we assessed the species‐specific detection space of AudioMoths, considering the influence of recording setup and location. ARUs had similar detection rates as a human observer when set up in locations with low vegetation. A further comparative analysis of bioacoustic and traditional roding surveys confirmed the comparability of the two methods. The last step comprised an evaluation of BirdNET as a means of automated species detection. Although BirdNET reliably detected Woodcock occurrence, our results underline the need for rigorous testing of computer‐aided data analysis as well as recording hardware before establishing PAM. We demonstrate that low‐cost open source ARUs and open access analysis tools can yield results on a par with traditional survey methods, and we provide specific recommendations for PAM for Eurasian Woodcock.

Effects of Urban Lighting at Night on Daytime Songbird Aggression

Many animals require sleep for daily functions. In spring, songbirds expend a great deal of energy early in the day by singing to attract mates and aggressively protecting territory. Some evidence suggests that artificial lighting at night (ALAN) in the environment of wild birds increases their nighttime activity that, in turn, contributes to sleep loss. This study investigated how ALAN affects northern cardinal (Cardinalis cardinalis) response to experimental aggressive intrusion by conspecifics. The hypothesis tested is that high ALAN near cardinal roosting spots would cause those individuals to have less energy for daytime aggressive behaviors. After selecting known dark and bright urban roosting locations, ALAN was measured at specific sites with active cardinal territories during nighttime hours. Observers returned in the morning to present cardinal territorial songs using standardized call playback protocols. Aggressive behaviors of responding cardinals were recorded and used to quantify aggression intensity. Standard aggression measures included response latency, distance from bird to speaker, number of songs, hops, and flip flops. An aggression index was obtained using factor analysis which confirmed that aggressive responses featured short latency and approach distances, few chips, and many songs, flip flops, and hops. Analysis of variance revealed that cardinals roosting in high ALAN engaged in significantly more aggressive morning responses; directly contradicting predictions and previously published findings. One possible explanation for this result is that high ALAN might improve sleep in cardinals. To test this, autonomous recording units (ARUs) were placed in study locations to record nighttime calling behaviors in high and low ALAN. Insufficient data were obtained to clarify whether nighttime activity was reduced significantly by ALAN exposure. However, results showed that nighttime activity by cardinals is strongly influenced by moonlight, and that determination of the influence of ALAN on nighttime and daytime avian behavior must include measures of lunar illumination.

Anuran occupancy varies with stream characteristics and flow across Arizona wilderness areas

Abstract Riparian ecosystems comprise less than 2% of the landscape in the arid western United States, yet they provide habitat and resources to over half of arid‐land wildlife species, including a broad diversity of anurans (frogs and toads). Despite this, relatively few studies describe riparian herpetofaunal communities in the Southwest, and remote wilderness areas are particularly undersurveyed. We employed passive acoustic monitoring to capture anuran advertisement calls in wilderness area tributaries of the Verde River, Arizona, USA. In the spring and summer of 2021 and 2022, 13–29 autonomous recording units (ARUs) were deployed along perennial, intermittent and ephemeral tributary stream reaches. We characterized sites based on the per cent of pool, riffle, run and side channel habitat within 100 m of each ARU and quantified substrate, wetted channel width and canopy cover. To relate anuran occupancy to environmental and hydrological variables, we evaluated acoustic detection data using single‐species, single‐season occupancy models. Four species were detected in this study: canyon treefrog (Hyla arenicolor), red‐spotted toad (Anaxyrus punctatus), Woodhouse's toad (Anaxyrus woodhousii) and non‐native American bullfrog (Lithobates catesbeianus), with canyon treefrog being the most ubiquitous species observed. Vocal activity of canyon treefrog was higher at perennial and intermittent sites compared to ephemeral sites, and the presence of pools was most strongly associated with canyon treefrog occupancy. Persistent, slow‐moving water best predicted the presence of breeding canyon treefrogs. Notably, this study did not detect several species with historical records in the middle Verde River catchment, including Arizona toad (Anaxyrus microscaphus) and northern leopard frog (Lithobates pipiens). Given climate change‐related flow declines and intensifying demands for water in arid lands globally, maintaining stream flows that provide consistent and suitable hydroregimes for anuran breeding and larval development is of increasing importance. Determining habitat use and flow regimes necessary to support anuran populations can aid in prioritization of conservation actions related to water management and predict how changes in water availability may impact stream‐breeding anurans.

Using autonomous recording units to identify and monitor western yellow‐billed cuckoo habitat

AbstractAutonomous recording units (ARUs) paired with signal classification software can be used to detect species‐specific calls, making them useful for evaluating patterns in avian occurrence and activity. However, classification of target signals is not always reliable and may be especially challenging for species that vocalize infrequently. We assessed the use of ARUs to identify and monitor habitat for a cryptic and federally threatened distinct population segment, the western yellow‐billed cuckoo (Coccyzus americanus) in mountainous xeroriparian drainages. Using Kaleidoscope Pro, we developed a call‐classifier and processed acoustic data collected in sites also surveyed using traditional human‐observer methods, applying the same spatial and temporal detection criteria to estimate breeding territories for each method. The classifier detected a total of 4,061 true positive calls at 4 sites, had an overall precision score of 0.07, recall score of 0.09, and F‐score (beta = 1) of 0.08, indicating high false positive and false negative classification rates. Our results were, however, consistent with other ARU studies of rare and cryptic species and ARUs estimated occupancy as effectively as human surveys with as little as 2 hours of daily recording. Total detections varied among sites, likely due to differences in cuckoo population densities and the interaction between topography and ARU detection space. Our results suggest that despite performance shortcomings of call‐classifiers, ARUs can be effective for monitoring cuckoos, with potential for providing higher resolution temporal and spatial information on activity and habitat use and may be particularly effective in remote locations where cuckoos often occur.

Comparing detection accuracy of mountain chickadee (Poecile gambeli) song by two deep-learning algorithms

The use of autonomous recording units (ARUs) has become an increasingly popular and powerful method of data collection for biological monitoring in recent years. However, the large-scale recordings collected using these devices are often nearly impossible for human analysts to parse through, as they require copious amounts of time and resources. Automated recognition techniques have allowed for quick and efficient analysis of these recordings, and machine learning (ML) approaches, such as deep learning, have greatly improved recognition robustness and accuracy. We evaluated the performance of two deep-learning algorithms: 1. our own custom convolutional neural network (CNN) detector (specialist approach) and 2. BirdNET, a publicly available detector capable of identifying over 6,000 bird species (generalist approach). We used audio recordings of mountain chickadees (Poecile gambeli) collected from ARUs and directional microphones in the field as our test stimulus set, with our custom detector trained to identify mountain chickadee songs. Using confidence thresholds of 0.6 for both detectors, we found that our custom CNN detector yielded higher detection compared to BirdNET. Given both ML approaches are significantly faster than a human detector and the custom CNN detector is highly accurate, we hope that our findings encourage bioacoustics practitioners to develop custom solutions for targeted species identification, especially given the availability of open-source toolboxes such as Koogu.

Performance of unmarked abundance models with data from machine‐learning classification of passive acoustic recordings

AbstractThe ability to conduct cost‐effective wildlife monitoring at scale is rapidly increasing due to the availability of inexpensive autonomous recording units (ARUs) and automated species recognition, presenting a variety of advantages over human‐based surveys. However, estimating abundance with such data collection techniques remains challenging because most abundance models require data that are difficult for low‐cost monoaural ARUs to gather (e.g., counts of individuals, distance to individuals), especially when using the output of automated species recognition. Statistical models that do not require counting or measuring distances to target individuals in combination with low‐cost ARUs provide a promising way of obtaining abundance estimates for large‐scale wildlife monitoring projects but remain untested. We present a case study using avian field data collected in the forests of Pennsylvania during the spring of 2020 and 2021 using both traditional point counts and passive acoustic monitoring at the same locations. We tested the ability of the Royle–Nichols and time‐to‐detection models to estimate the abundance of two species from detection histories generated by applying a machine‐learning classifier to ARU‐gathered data. We compared abundance estimates from these models with estimates from the same models fit using point‐count data and to two additional models appropriate for point counts, the N‐mixture model and distance models. We found that the Royle–Nichols and time‐to‐detection models can be used with ARU data to produce abundance estimates similar to those generated by a point‐count‐based study but with greater precision. ARU‐based models produced confidence or credible intervals that were on average 31.9% (±11.9 SE) smaller than their point‐count counterpart. Our findings were consistent across two species with differing relative abundance and habitat use patterns. The higher precision of models fit using ARU data is likely due to higher cumulative detection probability, which itself may be the result of greater survey effort using ARUs and machine‐learning classifiers to sample significantly more time for focal species at any given point. Our results provide preliminary support for the use of ARUs in abundance‐based study applications, and thus may afford researchers a better understanding of habitat quality and population trends, while allowing them to make more informed conservation recommendations and actions.

A large-scale assessment of eastern whip-poor-will (Antrostomus vociferus) occupancy across a gradient of forest management intensity using autonomous recording units

Conservationists spend considerable resources to create and enhance wildlife habitat. Monitoring how species respond to these efforts helps managers allocate limited resources. However, monitoring efforts often encounter logistical challenges that are exacerbated as geographic extent increases. We used autonomous recording units (ARUs) and automated acoustic classification to mitigate the challenges of assessing Eastern Whip-poor-will (Antrostomus vociferus) response to forest management across the eastern USA. We deployed 1263 ARUs in forests with varying degrees of management intensity. Recordings were processed using an automated classifier and the resulting detection data were used to assess occupancy. Whip-poor-wills were detected at 401 survey locations. Across our study region, whip-poor-will occupancy decreased with latitude and elevation. At the landscape scale, occupancy decreased with the amount of impervious cover, increased with herbaceous cover and oak and evergreen forests, and exhibited a quadratic relationship with the amount of shrub-scrub cover. At the site-level, occupancy was negatively associated with basal area and brambles (Rubus spp.) and exhibited a quadratic relationship with woody stem density. Implementation of practices that create and sustain a mosaic of forest age classes and a diverse range of canopy closure within oak (Quercus spp.) dominated landscapes will have the highest probability of hosting whip-poor-wills. The use of ARUs and a machine learning classifier helped overcome challenges associated with monitoring a nocturnal species with a short survey window across a large spatial extent. Future monitoring efforts that combine ARU-based protocols and mappable fine-resolution structural vegetation data would likely further advance our understanding of whip-poor-will response to forest management.

Male mating season range expansion results from an increase in scale of daily movements for a polygynous-promiscuous bird.

Males of species with promiscuous mating systems are commonly observed to use larger ranges during the mating season relative to non-mating seasons, which is often attributed to a change in movements related to reproductive activities. However, few studies link seasonal range sizes to variation in daily space use patterns to provide insight into the behavioral mechanisms underlying mating season range expansion. We studied 20 GPS-tagged male wild turkeys (Meleagris gallopavo), a large upland gamebird, during the mating and summer non-mating seasons to test the hypothesis that larger mating season ranges resulted from male wild turkeys expanding the scale of daily movement activities to locate and court females. We delineated mating and non-mating seasons based on intensity of gobbling, a vocalization tied to courtship behavior, recorded by autonomous recording units distributed across the study area. Mating season ranges were significantly larger than non-mating season ranges. Daily ranges were larger in the mating season, as were distances between roost sites used on consecutive nights. Variance in daily range size was greater in the mating season, but low temporal autocorrelation suggested considerable daily variability in both seasons. We found no evidence that male wild turkeys changed how they distributed daily movements within seasonal ranges, or differences in habitat use, suggesting larger mating season ranges result from male wild turkeys increasing the scale of their daily movements, rather than a systematic shift to a nomadic movement strategy. Likely, the distribution of females is more dynamic and ephemeral compared to other resources, prompting males to traverse larger daily ranges during the mating season to locate and court females. Our work illustrates the utility of using daily movement to understand the behavioral process underlying larger space use patterns.

Efficient quality assurance and quality control for passive acoustic monitoring data: reducing and documenting false-positive and false-negative errors

ABSTRACT Autonomous Recording Units (ARUs) are widely used to survey for a variety of taxa. This survey method allows for high spatial and temporal coverage but will typically include identification errors that can bias estimates of occupancy. In some instances, verifying all individual detections is prohibitive. To direct verification effort, we developed a model to estimate the probability that transcribers would agree on an identification. Agreement probability was positively influenced by transcriber skill, identification confidence, species commonness and some song types. In contrast, agreement probability was lower when an acoustic signal was classified as a trill. We evaluated our model on independent data where all species detections were verified, and verification effort (time) was quantified. Our model performed well at predicting transcriber agreement on independent data (AUC = 0.71). We applied the model to randomised subsets of the independent data to compare the cost benefit of three approaches to verification under varying effort. We show how modelling probability of transcriber agreement can be used to more efficiently direct verification of species acoustic tags. Our approach could be adapted elsewhere to quantify and reduce species misidentifications in unverified passive acoustic monitoring data for either manual processing or detections from automated classifiers.

Field tests of small autonomous recording units: an evaluation of in-person versus automated point counts and a comparison of recording quality

ABSTRACT The proliferation of small autonomous recorders makes it easier than ever to sample terrestrial acoustic animals and soundscapes. I conducted a comparison of four small recorders to evaluate their performance in a field setting: Wildlife Acoustics Song Meter Mini; Wildlife Acoustics Song Meter Micro; Open Acoustics Audiomoth; and Cornell SwiftOne. I address two questions: (1) How do in-person point counts compare to recorder-based point counts using these small autonomous recorders? (2) How does the quality of the recordings compare across these small autonomous recorders? To evaluate the performance of the recorders in point counts, I conducted in-person and recording-based point counts at ten locations. Each of the recorders performed similarly well at point counts, producing comparable estimates of species richness, although all of the autonomous recorders under-estimated species richness. To evaluate recording quality, I conducted a sound transmission test, broadcasting and re-recording sounds. Recorders varied in their frequency response above 12 kHz, but showed only subtle differences in the frequency response at frequencies below 12 kHz. I conclude that each of these types of small recorders provide bioacousticians with useful tools for conducting point counts, and for passive monitoring of animal sounds, with only subtle differences across the investigated models.

Calling for the future of conservation: a protocol for passive acoustic monitoring of small arboreal primates

Biodiversity conservation faces challenges due to a lack of accurate information on species occurrence. Various techniques have been used to survey species diversity and estimate population density, but monitoring species over large spatial and temporal scales remains challenging. Passive acoustic monitoring (PAM) has emerged as a cost-effective and non-invasive method for monitoring biodiversity. PAM utilizes autonomous recording units (ARUs) installed in different areas and is particularly relevant for monitoring threatened species in tropical forest regions. In the case of non-human primates, PAM has proven effective in detecting endangered species, monitoring populations, studying vocal behavior, and evaluating territory use. The black lion tamarin (Leontopithecus chrysopygus), an endangered species of the Atlantic Forest, relies on acoustic signals for communication. This study proposes a PAM protocol for monitoring arboreal primates using the black lion tamarin as a model. It reviews PAM's use in primate research and emphasizes defining target vocalizations. We recommend optimal recording conditions, including distance, height, and equipment. Recordings should be positioned high above the ground, considering the arboreal nature of primates. The choice of spatial distribution, including random placement, transects, and grids, depends on the research question and objectives. Lastly, the study addresses the recording schedule, considering periods of greater species activity, such as from sunrise to sunset. In summary, this study highlights PAM's potential for monitoring arboreal primates providing recommendations for vocalizations, recording conditions, equipment, spatial distribution, and schedules, contributing to effective monitoring, and supporting conservation efforts in tropical forests.

Mel-frequency cepstral coefficients outperform embeddings from pre-trained convolutional neural networks under noisy conditions for discrimination tasks of individual gibbons

Passive acoustic monitoring – an approach that utilizes autonomous acoustic recording units – allows for non-invasive monitoring of individuals, assuming that it is possible to acoustically distinguish individuals. However, identifying effective analytical approaches for individual identification remains a challenge. Our study investigates how the use of different feature representations impacts our ability to distinguish between individual female Northern grey gibbons (Hylobates funereus). We broadcast pre-recorded calls from twelve gibbon females and re-recorded the calls at varying distances (directly under the tree to ∼400 m away) using autonomous recording units. We evaluated the effectiveness of using different automated feature extraction approaches to classify gibbon calls: Mel-frequency cepstral coefficients (MFCCs), embeddings from three pre-trained neural networks (BirdNET, VGGish, and Wav2Vec2), and four commonly used acoustic indices. We used a supervised classification approach (random forest) to classify calls to the respective female and compared two unsupervised clustering approaches (affinity propagation clustering and hierarchical density-based spatial clustering) to evaluate which features were most effective for distinguishing female calls without using class labels. We used MFCCs as a baseline as previous work has shown they can be used to distinguish high-quality calls of individual gibbon females. Human annotators could only identify calls in spectrograms from recordings <350 m from the playback speaker with signal-to-noise ratio ∼ 0 dB, so our results focus on these recordings. Using supervised classification, our results confirmed the efficiency of MFCCs and the use of embeddings from one neural network (BirdNET) for effective acoustic classification of gibbon individuals at closer recording distances (signal-to-noise ratio > 10 dB), while the remaining features did not perform well. Contrary to our expectations, we found that MFCCs outperformed all other features for the unsupervised clustering tasks at closer distances and none of the features performed well at farther distances. The ability to acoustically discriminate animals under noisy conditions and from low signal-to-noise ratio calls has important implications for monitoring populations of endangered animals, such as gibbons. Focusing only on high signal-to-noise ratio calls for individual discrimination may not be possible for rare sounds, and future work should focus on developing effective approaches of feature extraction that can perform well across noisy, real-world conditions with a limited number of training samples.

Evaluation of an autonomous acoustic surveying technique for grassland bird communities in Nebraska.

Monitoring trends in wildlife communities is integral to making informed land management decisions and applying conservation strategies. Birds inhabit most niches in every environment and because of this they are widely accepted as an indicator species for environmental health. Traditionally, point counts are the common method to survey bird populations, however, passive acoustic monitoring approaches using autonomous recording units have been shown to be cost-effective alternatives to point count surveys. Advancements in automatic acoustic classification technologies, such as BirdNET, can aid in these efforts by quickly processing large volumes of acoustic recordings to identify bird species. While the utility of BirdNET has been demonstrated in several applications, there is little understanding of its effectiveness in surveying declining grassland birds. We conducted a study to evaluate the performance of BirdNET to survey grassland bird communities in Nebraska by comparing this automated approach to point count surveys. We deployed ten autonomous recording units from March through September 2022: five recorders in row-crop fields and five recorders in perennial grassland fields. During this study period, we visited each site three times to conduct point count surveys. We compared focal grassland bird species richness between point count surveys and the autonomous recording units at two different temporal scales and at six different confidence thresholds. Total species richness (focal and non-focal) for both methods was also compared at five different confidence thresholds using species accumulation curves. The results from this study demonstrate the usefulness of BirdNET at estimating long-term grassland bird species richness at default confidence scores, however, obtaining accurate abundance estimates for uncommon bird species may require validation with traditional methods.

Springtime bird use of agricultural landscapes examined by using autonomous recording units

Springtime bird use of agricultural landscapes examined by using autonomous recording units

Mocking Bird ( Mimus polyglottos ) calls potentially confound acoustic indices of bird diversity and provide a potential heuristic to distinguish them.

This study explores the increasing use of autonomous recording units (ARUs) in wildlife surveys. While ARUs offer cost-effective and efficient data collection, challenges arise in analyzing large datasets and accurately assessing species abundance. Our research focuses on avian communities, emphasizing the impact of vocal mimicry by Northern Mockingbirds ( Mimus polyglottos ) on survey accuracy. Utilizing the Merlin Bird ID application, we found an average accuracy rate of ~81.3%, with mockingbirds contributing ~31% of false positive identifications. Finding potential solutions for distinguishing mimics in bioacoustic survey data is crucial for enhancing accuracy as researchers increasingly adopt this methodology in the future.

Using autonomous recording units for vocal individuality: insights from Barred Owl identification

Using autonomous recording units for vocal individuality: insights from Barred Owl identification

Using bioacoustics to enhance the efficiency of spotted owl surveys and facilitate forest restoration

AbstractThe California spotted owl (Strix occidentalis occidentalis) is an older‐forest associated species that resides at the center of forest management planning in the Sierra Nevada and Southern California, USA, which are experiencing increasingly large and severe wildfires and drought‐related tree mortality. We leveraged advances in passive acoustic survey technologies to develop an acoustically assisted survey design that could increase the efficiency and effectiveness of project‐level surveys for spotted owls, allowing surveys to be completed in a single year instead of in multiple years. We deployed an array of autonomous recording units (ARUs) across a landscape and identified spotted owl vocalizations in the resulting audio using BirdNET. We then evaluated spatio‐temporal patterns in spotted owl vocalizations near occupied territories and the ability of a crew naïve to the location of occupied territories to locate spotted owls based on patterns of acoustic detections. After only 3 weeks of acoustic surveys, ≥1 ARU within 750 m of all 17 occupied territories obtained spotted owl detections across ≥2 nights. When active surveys using broadcast calling were conducted near ARUs with spotted owl detections by surveyors naïve to territory occupancy status and locations, surveyors located owls in 93% to 100% of occupied territories with ≤3 surveys. To further improve the efficiency of spotted owl surveys, we developed a statistical model to identify and prioritize areas across the Sierra Nevada for different survey methods (active only, acoustically assisted, no surveys) based on the expected probability of occupancy predicted from remotely sensed measurements of tree height and historical occupancy. Depending on managers' tolerance for false negatives, this model could help identify large areas that might not benefit from surveys based on low expected occupancy probabilities and areas where acoustically assisted surveys might enhance survey effectiveness and efficiency. Collectively, these findings can help managers streamline the survey process and thus increase the pace of forest restoration while minimizing potential near‐term adverse effects on California spotted owls.

Simultaneous passive acoustic monitoring uncovers evidence of potentially overlooked temporal variation in an Amazonian bird community

The vocal activity and detectability of tropical birds are subject to high levels of temporal heterogeneity, but quantifying patterns of diel and day‐to‐day variation in complex systems is challenging with traditional point count methods. As a result, research concerning stochastic temporal effects on tropical bird assemblages is limited, typically offering only broad conclusions, such as that overall activity is highest in the first few hours of the morning and some species are active at different times of the day. Passive acoustic monitoring introduces several advantages for studying temporal variation, particularly by enabling simultaneous and continuous data collection across adjacent sites. Here, we employed autonomous recording units to quantify temporal variation in bird vocal activity and observed species richness at an Amazonian reserve in Madre de Dios, Peru, a region featuring some of Earth's richest, most complex bird assemblages. We manually annotated 18 dawn hour recordings, collected simultaneously from three separate days at the same six sites, which represent various microhabitats and bird community compositions. We documented significant and consistent temporal variation in avian vocal activity levels and observed species richness within the dawn hour and across days. We found that temporal effects were stronger for vocal activity than for observed species richness and that vocal activity patterns over the course of the dawn hour varied between species. Our results indicate that overlooked temporal variation in Amazonian soundscapes may obfuscate the results of surveys that do not sufficiently account for temporal variables with simultaneous monitoring. While manual analysis of large volumes of soundscape data remains challenging, such data should be collected to supplement traditional surveys whenever possible. Rapid advances in the automated processing of acoustic data could lead to more efficient methods for reducing temporal bias and improving the calibration and accuracy of tropical bird surveys.

Software-dependent biases in the recognition of di- and tri-syllabic bird songs can create false interpretations of bird abundance and singing activity

The use of autonomous recording units (ARUs) for passive acoustic monitoring has recently gained a lot of importance. It is now widely used for scientific purposes and has been increasingly applied in the field of conservation biology. However, while ARUs may greatly increase cost-effectiveness in monitoring, they are not exempt from possible biases. In this study, we compared the performances of three different software in recognizing bi- and tri-syllabic bird songs and tested if their results varied significantly in a real setting. We focused on the interpretation of the relative abundance among habitats and the time of vocal activity, also using different numbers of training records to assess if the results of the software improved. Even with the quality measures of the recognizer being consistent with those of previously published studies, in several cases the results produced significantly different interpretations depending on the software used. While increasing the number of training records slightly improved the overlap in the estimates of activity among the different software, differences in the relative abundance of each song were present even with a fairly solid number of training records. Our results suggest that reliable biological interpretations might only be attained with a large number of training records and very high values of precision, recall, and the F-score. In fact, even with very high precision and F-scores, lower values for recall led not only to statistically significant differences, but also to different interpretations of relative abundance and time of activity between the two best-performing software. Thus, great caution must be taken in interpreting the results of bioacoustics studies that use automated recognition, and standards of minimum quality should be created for recognizers in scientific and technical studies.

Optimizing detection of the bobwhite reproduction call using passive acoustic monitoring

AbstractNorthern bobwhite (Colinus virginianus) populations have declined across much of their range. In response to these declines, wildlife biologists and managers have increased survey efforts and tried to optimize detection and capitalize on technological advances to improve population estimates and cost‐effectiveness. Our objective was to determine how environmental conditions influence detection of the reproduction call, or whistle, of masked bobwhite (C. v. ridgwayi), an endangered subspecies of northern bobwhite, using autonomous recording units (ARUs). We estimated the call intensity of the masked bobwhite reproduction call as 112 ± 0.5 decibels (mean ± SE) at 10 cm. We then broadcasted 16,284 calls during 17 trials to compare manual and automated call detection in recordings collected with ARUs. We used these data to model detectability of a bobwhite reproduction call, for when the bird is present and available, as a function of distance and weather conditions using generalized linear mixed models with trial as a random effect. Regardless of detection type, one model structure was competitive and suggested detection probability was a function of distance, wind speed, and wind direction. Detectability decreased with increased distance and wind speed and was influenced by wind direction. We demonstrate the use of our results to predict the probability of detecting a reproduction call during ARU‐based monitoring efforts. By understanding the effects of environmental factors on the detection of a bobwhite reproductive call, bobwhite surveys can be improved.